High-activity CAS researcher building benchmarks and datasets for full-duplex spoken dialogue
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
37
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-15) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.
Research profile
Output—
The output axis is not yet scored: this snapshot supplies no verified per-year publication list or total volume for this researcher. It would measure cumulative and per-year publication counts relative to field norms, rather than citations or influence.
Impact—
The impact axis is not yet scored because verified citation totals and field-normalized citation metrics are absent from this snapshot. It would measure the citations and scholarly influence accumulated by the published record relative to field norms.
Focus0/100
The focus axis scores 0. The field distribution is split evenly (0.5/0.5) between two near-duplicate spoken-dialogue labels, while the topic profile is dominated by Aquatic Ecosystems and Phytoplankton Dynamics (79) and Marine and coastal ecosystems (69), with Water Quality Monitoring (31 and 28) and Remote-Sensing Image Classification (13) behind. This incoherent signature drives the low score and most likely reflects namesake mixing in the OpenAlex record.
Collaboration60/100
Collaboration scores 60 from a raw count of 19 distinct co-authors. The recurring partner list includes Qingxiang Guo, Wenke Fan, Hongwei Cai, Weixu Wang, Shuofeng Zhao, and Dawei Yang, and both catalog papers place this researcher in a middle author position without corresponding-author status. The pattern suggests steady participation in team-written benchmark and dataset papers rather than a small fixed duo or a large solo-led network.
Activity97/100
Activity scores 97, the strongest axis in this profile. The year-by-year counts show continuous output from 2015 (1 work) through a 2021 peak of 21, then 13 works in 2022, 12 in 2023, 6 in 2024, 7 in 2025, and 14 already recorded for 2026, with 27 works in the last three years. The 2026 figure approaching the historical peak suggests currently accelerating output.
Reproducibility38/100
Reproducibility averages 38 across the 2 papers Evolve has reviewed. ECHO is rated medium: its matched-insertion design is reusable, but the verdict flags label-dependent interruption acoustic processing and a missing gain-free or waveform-reuse control. DuplexDrama is rated low, mainly because the 800-hour synthesized corpus lacks human perceptual validation. Both are arXiv releases, so openly shared code and data would lift this axis.
Research summary
Yingxin Shang (Chinese Academy of Sciences, CN) is profiled here from public bibliographic data, with two 2026 arXiv papers in the Evolve catalog. The recent work sits in speech processing and spoken dialogue systems: ECHO introduces a matched-contrast benchmark for context-sensitive turn-taking in full-duplex dialogue, and DuplexDrama presents a synthesized 800-hour full-duplex dialogue dataset with personas, expressive speech, and sound events. The bibliometric snapshot shows a highly active record — activity is scored 97, with 27 works in the last three years, a 2021 peak of 21 works, and 14 already recorded for 2026 — alongside an h-index of 37 and an i10-index of 68. Collaboration is moderate (score 60, 19 distinct co-authors), and reproducibility readiness across the two reviewed papers averages 38 (medium for ECHO, low for DuplexDrama). The topic distribution is split between spoken-dialogue themes and aquatic-ecosystems/water-quality topics, so this profile should be read as name-resolved from OpenAlex rather than a verified single identity. Focus is scored 0, and the output and impact axes lack sufficient underlying data to score.
Collaboration pattern
Collaboration is scored 60 on the strength of 19 distinct co-authors. Recurring names include Qingxiang Guo, Wenke Fan, Hongwei Cai, Weixu Wang, Shuofeng Zhao, and Dawei Yang. Both reviewed papers carry a middle author position with no corresponding role, consistent with a contributing member of a bench-scale dialogue-systems team rather than a hub coordinating many groups. Co-author identities come from the OpenAlex snapshot and may include namesakes.
Research keywords
Highlights
ECHO: matched-contrast benchmark exposing YIELD bias in full-duplex turn-taking
DuplexDrama: 800-hour synthesized full-duplex dialogue dataset
Two 2026 releases targeting full-duplex evaluation infrastructure
Both catalog entries appeared in 2026 with this researcher in a middle author position and no corresponding role, and both target evaluation infrastructure for full-duplex spoken dialogue. Evolve rates their reproducibility readiness at a mean of 38 (medium and low), leaving room for open code, controls, and human validation.
Timeline
- 2015Earliest year in the OpenAlex work-count record, with 1 work recorded.
- 2018Output reached 10 works, an early ramp-up in the yearly record.
- 2021Peak year of the record, with 21 works recorded.
- 202614 works recorded to date; the Evolve catalog adds two 2026 arXiv papers on full-duplex spoken dialogue (ECHO and DuplexDrama).
Papers in the Evolve catalog
Frequent co-authors
About this profile
This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.
- Bibliometric scores and raw counts come from a single OpenAlex snapshot (identity: openalex) and are not independently verified.
- The topic profile (aquatic ecosystems, water quality, remote sensing) diverges sharply from the reviewed spoken-dialogue papers, so the snapshot may aggregate namesakes; this profile is name-resolved from public bibliographic data, not a verified biography.
- Only 2 Evolve-reviewed papers (both 2026 arXiv, middle authorship) anchor the qualitative claims; career facts, awards, and a full publication list are out of scope.
- Citation totals and per-paper citation data are absent, so the output and impact radar axes are left unscored.